J. Waechter is clinical associate professor, Department of Critical Care, Cumming School of Medicine, University of Calgary, Calgary, Alberta, Canada.
J. Allen is professor, Department of Medicine, University of North Dakota School of Medicine and Health Sciences, Grand Forks, North Dakota.
Acad Med. 2022 Oct 1;97(10):1484-1488. doi: 10.1097/ACM.0000000000004758. Epub 2022 May 24.
Clinical reasoning is a core competency for physicians and also a common source of errors, driving high rates of misdiagnoses and patient harm. Efforts to provide training in and assessment of clinical reasoning skills have proven challenging because they are either labor- and resource-prohibitive or lack important data relevant to clinical reasoning. The authors report on the creation and use of online simulation cases to train and assess clinical reasoning skills among medical students.
Using an online library of simulation cases, they collected data relevant to the creation of the differential diagnosis, analysis of the history and physical exam, diagnostic justification, ordering tests; interpreting tests, and ranking of the most probable diagnosis. These data were compared with an expert-created scorecard, and detailed quantitative and qualitative feedback were generated and provided to the learners and instructors.
Following an initial pilot study to troubleshoot the software, the authors conducted a second pilot study in which 2 instructors developed and provided 6 cases to 75 second-year medical students. The students completed 376 cases (average 5.0 cases per student), generating more than 40,200 data points that the software analyzed to inform individual learner formative feedback relevant to clinical reasoning skills. The instructors reported that the workload was acceptable and sustainable.
The authors are actively expanding the library of clinical cases and providing more students and schools with formative feedback in clinical reasoning using our tool. Further, they have upgraded the software to identify and provide feedback on behaviors consistent with premature closure, anchoring, and confirmation biases. They are currently collecting and analyzing additional data using the same software to inform validation and psychometric outcomes for future publications.
临床推理是医生的核心能力，也是常见的错误来源，导致误诊率和患者伤害率居高不下。尽管人们努力提供临床推理技能的培训和评估，但由于其劳动和资源密集型，或者缺乏与临床推理相关的重要数据，这些努力都证明具有挑战性。作者报告了创建和使用在线模拟病例来培训和评估医学生的临床推理技能。
他们使用在线模拟病例库，收集与创建鉴别诊断、分析病史和体检、诊断理由、检查医嘱、解读检查以及对最可能诊断的排序相关的数据。这些数据与专家创建的记分卡进行了比较，并为学习者和教师生成了详细的定量和定性反馈。
在对软件进行初步试点研究以排除故障后，作者进行了第二项试点研究，其中 2 名教师开发并向 75 名二年级医学生提供了 6 个病例。学生完成了 376 个病例（平均每个学生 5.0 个病例），生成了超过 40200 个数据点，软件对这些数据点进行了分析，为与临床推理技能相关的个别学习者形成性反馈提供信息。教师报告说，工作量是可以接受和可持续的。
作者正在积极扩展临床病例库，并通过我们的工具为更多学生和学校提供临床推理的形成性反馈。此外，他们已经升级了软件，以识别和提供与过早封闭、锚定和确认偏差一致的行为的反馈。他们目前正在使用相同的软件收集和分析更多数据，以告知未来出版物的验证和心理测量结果。
